the so-called Vikramasamvat or, more correctly, the Samvat of the Mâlavesas, the lords of Mâlava, which began in 57 B. C. Rudradâman's inscription consequently dates from the year 21-22 A.D., and it is thus certain that the word Pahlava was used in India at the beginning of the first century A. D. These circumstances make it impossible to accept Professor Nöldeke's inferences from the occurrence of the softened Iranian forms. But the mere mention of the Pahlavas would show that Manu's verse cannot have been composed before the beginning of the first century B. C. The Parthian dynasty of the Arsacides was founded in the middle of the third century B. C., and its sixth ruler, Mithradates I, according to some classical authors, invaded India about the middle of the second century1. Coins of an Arsaces Theos and of an Arsaces Dikaios, who uses also the Prakrit language and the North-Indian alphabet, have been found in the Panjâb, and belong to the same or a little later times. As the Brâhmans are ever ready to give foreign nations, with which they come into contact, a place in their ethnological system, it is quite possible that about the beginning of the first century B. C. an Indian origin might have been invented for the Pahlavas. But even this reduction of the remoter limit of the Manu-smriti is, in my opinion, not quite safe. For though the evidence for the genuineness of Manu X, 43-44 is as complete as possible, and though the varia lectio for Pahlava, which Govinda offers, probably deserves no credit 3, there is yet a circumstance which raises a suspicion against the latter reading. Parallel passages, closely resembling Manu's two verses, are found in the Mahâbhârata XIII, 33, 21-23a and XIII, 35, 17–18, where the names of the degraded Kshatriya races are likewise enumerated, and the cause of their degradation is stated

1 Lassen, Indische Alterthumskunde, II2, 334.

2 Sallet, Die Nachfalger Alexanders des Gr. pp. 51, 156–157.

3 The commentators and MSS. all give the two verses. If some MSS. of Medhâtithi read Pahnava for Pahlava, that is a clerical mistake caused by the similarity of the subscribed Devanâgarî la and na. Govinda's var. lect. Pallava is improbable, because the other races mentioned in the second line of verse 44 all belong to the North of India, while the Pallavas are, as far as we know, confined to the South.

in exactly the same or similar words.

Both passages name

the Yavanas, and one also the Sakas. But neither mentions the Pahlavas. Hence it becomes doubtful if the original version of these Slokas really did contain the latter name. It is further not impossible that its insertion is not due to the first remodeller of the Mânava Dharma-sûtra, but has crept in later accidentally, in the place of some other name. The Indian Pandits are not strong in ethnology and history, and habitually careless with respect to the names of peoples and countries, which they frequently alter, or substitute in their works one for the other. I have, therefore, not the courage to reduce the terminus a quo by more than a hundred years on the strength of this single word, which occurs in a verse that evidently has had originally a different form. I think it safer to rely more on the mention of the Yavanas, Kâmbogas, and Sakas, and to fix the remoter limit of the work about the beginning of the second century A. D., or somewhat earlier.

This estimate of the age of the Bhrigu Samhitâ, according to which it certainly existed in the second century A. D., and seems to have been composed between that date and the second century B. C., agrees very closely with the views. of Professor Cowell1 and Mr. Talboys Wheeler 2. It differs considerably from that lately expressed by Professor Max Müller, who considers our Manu to be later than the fourth century 3, apparently because a passage quoted from Vriddha Manu, which he takes to be a predecessor of our Samhitâ, mentions the twelve signs of the zodiac. I do not think that it has been proved that every work which enumerates the râsis must be later than the period when Ptolemy's astronomy and astrology were introduced into India. But irrespective of this objection, Professor Max Müller's opinion seems to me untenable, because, according to Professor Jolly's and my own researches, the Vriddha or Brihat Manu, quoted in the digests and commentaries, is not earlier, but later than Bhrigu's Samhitâ. Whatever may be thought

1 Elphinstone, History of India, p. 249 (edition of 1874).

2 History of India, vol. ii, p. 422.

3 India, what can it teach us? p. 366.

See above, p. xcvii.

of the details of my inferences and conclusions, I believe that the rudimentary state of the legal theories in our Samhitâ, as compared with Yâgñavalkya and Nârada (fourth or fifth century A. D.), the fact that the Brihaspati-smriti of the sixth or seventh century A. D. was a Vârttika on our text, and the assertion of Medhâtithi, that he knew in the ninth century commentaries belonging to a remote antiquity, force us to place it considerably before the term mentioned by Professor Max Müller.

III.

It now remains to give an account of the materials on which my translation is based, and of the manner in which they have been used. Among Sanskrit works the commentaries of Medhâtithi, Govindarâga, Sarvagña-Nârâyana, Kullûkabhatta, Râghavânanda, and Nandanâkârya, as well as an anonymous Tippana, contained in a Kasmîr MS. of the Manu-samhitâ, are the sources on which I have chiefly relied. Among the earlier translations, Sir William Jones' famous versio princeps and Professor J. Jolly's annotated German translation of chapter VIII and chapter IX, 1102 have been carefully used. Occasionally Mr. Loiseleur Deslongchamps' well-known edition of the text, the English translation of chapters I-III, 33 by Târâkand Kakravartî (Kuckerbutty) 2, and the Mârâthî translation of Ganârdan Vâsudev Gurgar3 have been consulted. Sir G. C. Haughton's edition and various Indian reprints of the text have been left aside, because they mostly repeat Kullûka's readings or give variae lectiones for which no sufficient. authority is shown.

Among the Sanskrit commentaries on the Manu-smriti the oldest extant is the voluminous Manubhâshya of Bhatta

1 Published in the Zeitschrift für vergleichende Rechtswissenschaft, vol. iii. 2 I have used the copy of the India Office Library, 19-27, 17. The name of the author is given by Professor Goldstücker, On the Deficiencies, &c., p. 5,

note.

3 Published with the text of Manu, at the Nirnayasagar Press, Bombay, 1877.

Medhâtithi, the son of Bhatta Vîrasvâmin. As its title, bhâshya, indicates, it is not a gloss which paraphrases every word of the text. Its aim is to show the general sense of Manu's dicta, to elucidate all really difficult passages, and to settle all doubtful points by a full discussion of the various possible interpretations, and of the opinions advanced by others. In carrying out this plan Medhâtithi displays a great amount of learning and not inconsiderable ability. He carefully uses a number of more ancient commentaries on Manu, and shows a full acquaintance with the Sâstras requisite for the successful explanation of his text, with Vedic literature, grammar, Mîmâmsâ, the Dharmasûtras1 and other Smritis, Vedânta, and the Mahâbharata. At the same time he avoids the common fault of Sanskrit commentators,-an undue copiousness in quotations which bear only remotely on the subject under consideration. Moreover, he frequently enhances the value of his explanations by illustrating Manu's rules by instances taken from every-day life, a point which most Hindu writers on law and on kindred subjects entirely neglect. Finally, he frequently takes up a much more independent position towards his author than the other commentators dare to assume. Thus he does not shrink from declaring that many verses are arthavâdas, without legal force, and that many single words have been inserted merely vrittapûranârtham, ‘in order to make up the verse.' His chief weakness, on the other hand, which is not unfrequently observable, and which has drawn on him Kullûka's stricture 2 that he brings forward 'both valuable and valueless remarks, consists in a disinclination to decide between conflicting interpretations and in his sometimes placing side

1 Medhâtithi quotes the Dharma-sûtras in general, and Gautama, Baudhâyana, Âpastamba, and Vasishtha, as well as some other lost works, in particular. Among the lost Dharma-sûtras which he used, is a Kâtyâyanîya-sûtra, quoted on Manu VIII, 215, which seems to have treated the civil law in detail, and probably is the original of the metrical Kâtyâyana-smriti, from which the digests give so numerous extracts.

2 See the concluding verses of Kullûka's commentary. Sir W. Jones' statement that Medhâtithi's work is reckoned 'prolix and unequal' (Preface to the Translation, p. xvii, St. Grady) is probably based on this remark of Kullûka.

by side, as equally admissible, widely divergent opinions. This vacillation is perhaps justified in a restricted number of passages, where the text is really ambiguous or very obscure. But more commonly it seems to be due solely to an excessive veneration for the views of his predecessors 1, whose commentaries, in part at least, possessed a high antiquity and a great reputation, or whom he had personal reasons to respect. On several occasions he mentions certain explanations as those of the Pûrvas or Kiramtanas, i. e. of the ancient commentators. Thus he remarks on Manu IV, 223,' But the exposition given above is the view of the Ancients; hence it has also been given by us 2.' In another case, when explaining Manu IX, 141 and 147, he notes that his interpretation is that of upâdhyâya, i.e. of his own teacher from whom he learnt the Manu-samhitâ. Disagreeable as this want of decision may be to those who look to a commentary for a concise and authoritative explanation of its text, yet it is not without advantages. His copiousness in quoting the opinions of his predecessors makes his work extremely important for the student of the history of the Manu-smriti and of the Hindu law. The Bhashya clearly proves that Manu's text had been made for centuries an object of deep research, and that many of its verses had given rise to widely different interpretations. It shows, further, that a good many various readings existed. Finally, a comparison of the later still extant commentaries leaves no doubt that these in general are based on the Manubhâshya, and that even their divergent opinions and readings are frequently derived from the earlier work. Under these circumstances the question of

1 Though the opinions of others' are mentioned very frequently, and though sometimes those of three or four predecessors are contrasted, Medhâtithi gives only once the name of an earlier commentator, Manu IX, 253, ajant काचित्फलश्रुतिः सा सर्वार्थवाद इति कावर (?) [v. 1. कार (?)] विष्णुस्वामी यदत्र तत्त्वं तद्द

The name seems to be Vishnusvâmin. But it

is uncertain what the corrupt word, preceding it, may hide.

2

यत्तु प्राग्व्याख्यातं तत्पूर्वेषां दर्शनमित्यस्माभिरपि वर्णितम् ॥ Compare also the remark on Manu V, 128, तत्र चिरंतनै व्याख्यातम् ॥